Survey sampling header

Surveying cheese rind microbial diversity

What happens when you collect 137 cheeses from around the world and sequence their DNA? You end up eating a ton of cheese. But more importantly, you learn a lot about what bacteria and fungi live in a typical cheese rind, that sea salt may be a previously unrecognized source of cheese microbes, and whether the notion of ‘microbial terroir’ holds up in artisan cheeses. In this Science Digested, I provide a digested version of recent research that used new DNA-sequencing approaches to broadly survey the microbial diversity of artisan cheese rinds.

Cheese Rind Communities Provide Tractable Systems for In Situ and In Vitro Studies of Microbial Diversity Benjamin E. Wolfe, Julie E. Button, Marcela Santarelli, and Rachel J. Dutton Cell, 2014, http://dx.doi.org/10.1016/j.cell.2014.05.041

Full disclosure: As is clear from the citation above, I helped lead this project while I was a postdoctoral researcher in Rachel Dutton’s lab at Harvard University. I am delighted to use our ‘Science Digested’ section of MicrobialFoods.org to explain the results of this study.

As bloomy, natural, and washed cheeses are aged, microbes colonize the cheese surface creating a rind. These cheese rinds are a community of different species of bacteria and fungi. By breaking down the cheese curd and creating colors and textures, these rinds create the distinct flavors and aesthetics that define an artisan cheese. Previous studies have described the diversity of bacteria and fungi living in these cheese rinds. But usually only one or two cheeses from one particular part of the world are covered in one study. This has limited our ability to understand how microbial diversity of cheese rinds varies across geographic regions, across cheese styles, and across various other production parameters (milk type, milk treatment, etc.).

Naturally aged cheeses develop three types of rinds on their surfaces: bloomy rinds (left), natural rinds (middle), washed rinds (right).

Naturally aged cheeses develop three types of rinds on their surfaces: bloomy rinds (left), natural rinds (middle), and washed rinds (right).

In an article recently published in Cell, we present the first large-scale survey of the microbial diversity of cheese rinds. By examining the diversity of bacteria and fungi living in the rinds of 137 cheeses made throughout the US and North America, we paint a comprehensive picture of what bacteria and fungi form ‘typical’ bloomy, natural, and washed rind cheeses. In addition to describing the diversity of microbes that are colonizing these artisan cheeses, we also demonstrate that these tractable microbial communities have the potential to be a ‘lab rat’ for understanding how microbial communities are formed. While the major driver behind this study is to advance basic biology, some of the findings are directly applicable to how cheese is made and aged. Below is a summary of some of the most important take-home messages that have practical implications for those who make or sell artisan cheeses, or for people who just enjoy knowing more about what they are eating.

Rinds with Similar Microbial Compositions Form in Different Parts of the World It’s widely believed that microbial foods, such as cheese or sourdough bread, will develop unique microbial communities in different geographic regions. This idea of geographically unique microbes in fermented foods, sometimes referred to as ‘microbial terroir,’ is based on the assumption that different parts of the world harbour unique microbial communities and these unique communities will colonize and help characterize the foods produced in a region. These region-specific communities may have certain properties which in turn will lead to the production of specific flavors or aesthetics. This very appealing notion of microbial terroir has become popularized (including by yours truly), but it has rarely been rigorously tested by science. Datasets that span large geographic regions are not widely available for most fermented foods so this has remained a largely unanswered question.

Our survey of 137 cheeses made across the United States and Europe begins to address this question of microbial terroir, and suggests that geographic region doesn’t necessarily play a huge role in defining the microbes that grow on cheese rinds. We used high-throughput (also known as “next-generation“) DNA sequencing, which allowed us to sequence the DNA of bacterial and fungi from many rind samples at one time. Our data allowed us to identify bacteria and fungi in rind communities to the level of genus.

This is the most important figure from our survey of rind microbial diversity. See text for a full explanation.

Bacterial and fungal diversity of 137 cheeses sampled around the world. The top row of columns shows data for bacteria. The bottom row shows data for fungi. The tree at the top clusters these data by how similar the microbial communities are in one cheese compared to another. The colored circles along the bottom of the figure indicate whether the cheeses were natural, washed, or bloomy rinds. Numbers indicate cheese IDs that can be used to get more metadata in the table referenced below. Click to zoom in and to download a higher resolution version. See text for a full explanation.

When we looked at patterns of diversity across the rinds sampled, we see that different geographic regions don’t have unique microbial communities. Communities with very similar compositions (with the same types of bacteria and fungi) can form on cheeses made in very different parts of the world. The converse is also true: a single cheese producer can produce different cheeses with very different microbial communities.

So if where you make a cheese doesn’t determine the diversity of the rind, what does? We found that the environment the cheesemaker creates on the surface best explains differences in rind diversity. As cheesemakers age cheeses, they create unique environments on the rind surface through techniques such as washing the rind surface to create a washed rind or inoculating the cheese with molds to create a bloomy rind. These different techniques for creating specific rind types best explained how diversity was distributed across all 137 different cheeses. This lack of correlation between geography and microbial communities has lead us to commonly describe rind development as an “if you build it, they will come” situation. The place where you make a cheese doesn’t have as much of an impact as the way you make a cheese. If a cheesemaker in France, a cheesemaker in Wisconsin, and a cheesemaker in California create approximately the same type of conditions in their aging environment on a similar type of cheese, there is a good chance they will develop very similar microbial communities in the rind.

It’s really important to note that the lack of correlation between geography and composition isn’t necessarily due to the diversity of microbes being swamped out by starter cultures added to the cheese production process. We see examples of raw milk cheeses where few or no starter cultures are added having similar compositions in very disparate parts of the world. It’s also very important to note that we haven’t completely disproven microbial terroir in cheese. First off, we were only viewing microbial diversity at the level of genus. We know from prior research that at the level of species or even strains, that there can be important variation that could translate to unique cheese properties. Our lack of correlation between the geographic distance between cheesemaking locations and the composition of cheese rinds doesn’t mean that individual cheesemakers don’t develop unique microbes on their cheeses. In fact, we did see that some cheesemakers within geographic regions could have similar styles of cheeses with different compositions.

Moisture is a Major Driver of Microbial Diversity on Cheese Rinds
It may be fairly intuitive that different approaches for creating a rind will lead to differences in rind microbial community composition. But what is the underlying mechanism creating those differences? What is the most important environmental variable that cheesemakers manipulate to create differences in rind communities? Our study found that moisture explains the most variation in community composition. Natural, washed, and bloomy rinds span a continuum of moisture, created by the environment where the cheese is aged ( how long the cheese is aged (longer = drier). Bloomy rinds are the wettest cheeses being the bloomy rinds that are aged in high humidity environments and aged for the shortest amount of time (leading to lower moisture loss).

What microbes show the greatest response to moisture? Bacteria such as Vibrio, Pseudoaltermonas, Psychrobacter, and Pseudomonas, as well as the fungus Galactomyces (more commonly known as Geotrichum) were all more abundant in wetter cheeses. The bacteria Brevibacterium and Brachybacterium, and Staphylococcus as well as the molds Scopulariopsis and Aspergillus, were all more abundant on drier cheeses, such as natural rind cheeses.

Bacteria Usually Found in Marine Environments are Abundant in Cheese Rinds
Some of the microbes that form cheese rinds come from the starter cultures cheesemakers purchase and add to their milk. Some genera detected in our survey, including Geotrichum, Brevibacterium, and Penicillium are inoculated into cheese. But a large amount of the rind microbial diversity comes from the raw milk (if a raw milk cheese) or environmental sources (cave environment, materials used for cheesemaking, etc.). One striking pattern in our study was the abundance of genera of bacteria that are usually found in the marine environment. These genera include Pseudoalteromonas, Halomonas, and Vibrio.

How do these marine bacteria get to the rinds of cheese? One potential source is the natural sea salt that cheese makers use in the production of cheeses. Previous work in California has found these marine bacteria in brines used in cheese production and research in Korea has shown that salt used for producing kimchi contains viable bacteria.

What might these marine bacteria contribute to cheese flavors and aromas? When we zoomed in on one Pseudoaltermononas species, we found some clear indicators that these marine bacteria can have huge impacts on cheese quality. The Pseudoaltermononas genomes contain genes that encode proteases and lipases that are adapted to work in the same temperature range that is used to age cheeses. Pseudoaltermononas also possesses the enzyme methionine gamma-lyase which is responsible for producing many of the sulfur compounds that give some cheeses their pungent odors. Previously, scientists were only aware of these enzymes in starter cultures such as Brevibacterium, so these data suggest that uninoculated microbes could have large impacts on cheese quality.

In Bacterial-Fungal Interactions, Fungi are the Major Players
It’s long been known that species of bacteria and fungi growing on cheese rinds are not passively hanging out together. Some interactions are positive, like the deacidification of cheese rinds by yeasts that promote the growth of ripening bacteria such as Brevibacterium. Some interactions are negative, like the inhibition of mold growth by certain yeasts. To better understand the type and strength of interactions across a wide number of bacterial-fungal combinations, we grew cultures of each bacterium and fungus in pairwise combinations and measured how well they grew alone vs. together.

Top figure shows bacterial responses to a fungal partner. Bottom figure shows fungal responses to bacterial partners. See text for detailed explanation.

Top figure shows bacterial responses to a fungal partner. Bottom figure shows fungal responses to bacterial partners. A green asterisk indicates a positive growth response that is statistically significant. A red asterisk indicates a negative growth response. See text for detailed explanation.

What might this mean for cheese makers who are trying to manage the development of a rind? It might mean that altering the composition of fungi that are inoculated into a cheese may have a greater impact on the rind community than altering the bacteria. It might also mean that if you want a certain bacterial species to increase in abundance, you may have to do that not just by increasing the inoculation of the bacterium, but also by altering the abundance of any fungi that the bacterium may interact with. There’s much more to explore here, including the mechanisms behind these interactions. Stay tuned for more as the Dutton and Wolfe labs continue to dissect the microbial diversity of these fascinating microbial communities.

 

There’s a whole lot more packed into this paper. I will cover the rest in future articles at MicrobialFoods.org, but I encourage you to check out the whole paper at Cell. If you can’t get behind their paywall, get in touch and I can send you a PDF of the paper via email. There is a useful supplemental table that you can find here that provides metadata for the survey data figure above.

Anything strike you as surprising in this paper? Or perhaps there is some insight that may inform how you think about making or buying cheese? Please share your thoughts in the comment section below!

There are 6 comments on this article

  • Great job Benjamin & Bronwen, this site is brilliant and approachable – well done!
    Andy

    Reply to this comment
  • Jim Wallace says:

    This work is brilliant and should begin to make many cheese makers step back and rethink how they go about the development of rinds.

    Reply to this comment
  • This site is just amazing. I haven’t made cheese in years, but I’m about to revisit this, and these articles are a source of inspiration.

    Please do e-mail me a PDF of the Cell article when you get a chance, and keep up the good work!

    Reply to this comment
  • Shawn says:

    As a microbiologist who works mostly with bacteria, I’m interested to learn more about fungi. In cheese, are bacteria or fungi more dominant? There is not likely a straightforward answer to that. In general, are fungi as diverse or more diverse compared to bacteria? Most microbiome studies focus on the bacterial diversity, but studies like yours and others are now looking at fungi. I’m curious if there are as many interesting unstudied fungi as there are bacteria out there? In soil samples, do microbes or fungi dominate?

    Reply to this comment
    • Hi Shawn –

      Thanks for visiting the website and for your question.

      It totally depends on the type of cheese. In Camembert, fungi are much more dominant. In a washed rind cheese like Limburger, bacteria are much more dominant. In general, bacteria are more diverse than fungi, but again that depends on the cheese.

      For soil, the same is true. Some systems are dominated by fungi (many forest soils) while other systems may be much more bacterial (salt marsh biofilms).

      I agree that we need more studies that incorporate both bacteria and fungi…. and other microbes (protists, archaea, etc.).

      Best,

      Ben

      Reply to this comment
  • Where did these other strains come from? There are a few ways besides starter cultures that microorganisms can enter the “microbial superhighway” of the cheese. Raw milk has plenty of microbes in it, as does the environment where the cheeses are aged, and many of the unexpected strains likely attached to the rind as the cheeses aged. After cheese has curdled but before it’s aged, cheesemakers also add salt to enhance the flavors and slow down the growth of bacteria. Several of the strains found in the cheeses are salt-tolerant ocean bacteria, which likely hitched a ride in the sea salt that the cheesemakers added at that step.

    Reply to this comment

Leave a Reply to Andy Swinscoe Cancel reply

Your email address will not be published. Required fields are marked *